Comment by Mr_N | Hacker News Reader

Comment by Mr_N | Hacker News Reader

Parent Full thread

Sorry if I was unclear; I meant to express that I had indeed quite some time ago come to the realization, but due to placement as I said and not due to any solid proof as is cited in the article.

Secondly; I'm not sure I completely see you as disagreeing with my stated end goal. We want people to realize they are driving a potential weapon, and I think that's a change that'll have to come in far more areas than just traffic enforcement (better education, assessment of readiness, etc), but even in this dream world where we've made those changes, I don't think it's unreasonable to assume we will still find places in which if drivers tend to exceed a certain speed, there comes a higher likelihood of said harm. And if someone is putting others at risk by violating this, it comes down to a question of how you respond to that, and I at least would pen a reasonable solution at trying to proactively influence their behavior patterns. (unless I totally misunderstood how you were approaching my statement and this just went off in a random direction...)

even in this dream world where we've made those changes, I don't think it's unreasonable to assume we will still find places in which if drivers tend to exceed a certain speed, there comes a higher likelihood of said harm

But the question is not whether people's behavior affects the likelihood of harm; of course it does, in all areas of life, not just driving. The question is what is the appropriate response. See below.

And if someone is putting others at risk by violating this, it comes down to a question of how you respond to that, and I at least would pen a reasonable solution at trying to proactively influence their behavior patterns

And you do that, IMO, by attaching punishments to actual harm, not to an increased likelihood of harm (which must always be an estimate only).

First of all, the legislator who drafts the speeding laws and the bureaucrat who sets the speed limits for particular roads can't possibly know as much as the driver who's actually on the spot about what a reasonably safe speed is in a particular situation; speeding laws can't adjust for varying conditions.

Second, if punishments are attached to actual harm, and fall on the person who causes the harm, then that person has an incentive to make accurate estimates of the risk of harm. The legislator and the bureaucrat have no such incentive; but if the law can only impose punishment when there has been actual harm, the legislator and the bureaucrat can only do a limited amount of damage by trying to game the criteria.

However, if punishments can be attached to just an estimated increase in risk of harm, rather than to actual harm, legislators and bureaucrats have an incentive to exaggerate the estimated risk of harm, since it allows them to lower the threshold for penalties and thereby bring in more revenue. But drivers themselves, who can see perfectly well that the criteria are being gamed, respond by trying to evade the law. The end result is that the costs of the system that is in place to catch speeders, and the escalating arms race between the speeders and the government, far outweigh any benefits that might accrue from imposing punishments when there has been no actual harm.

This does two things that I don't agree with though. One, assumes that people can accurately judge -anything- appropriately about what is safe in their surroundings. I'll trust someone so far as to not be too afraid to go outside lest someone decide to run me over with a crop thresher, but I've seen enough "No officer I'm not too drunk to drive" (with car wedged in a tree) to trust that judgement carte-blanche.

Secondly, it seems to ignore two things, first, you say that with limits, people game the system. I don't disagree. But without limits, I expect the exact same gaming, "I feel comfortable driving at 100mph on this road", but far more unbounded. Combined with some pretty reasonable evidence towards higher speeds increasing the risk of accidents (http://www-nrd.nhtsa.dot.gov/Pubs/97.835.PDF), one can see the outcome this would lead to.

I think that your interpretation of the report is incorrect. What that report shows is that fatal accidents are more likely on roads that have higher speed limits, which is quite a different thing from saying higher speeds caused the accidents.

And, in fact, there is evidence that people do in fact judge safe speeds appropriately. Look at report FHWA-RD-92-084 [1], which was sponsored by the Federal Highway Administration and concludes that raising speed limits did not have a significant effect on speeds of motorists. Lowering the speed also did not have a significant effect on speeds, with the result that people simply ignored the speed limit and drove what they thought was appropriate in both cases.

There was also evidence that lowering the speed limit did not, in fact, decrease accidents.

1. http://www.ibiblio.org/rdu/sl-irrel/index.html

One, assumes that people can accurately judge -anything- appropriately about what is safe in their surroundings.

I didn't assume that; I only assumed that the people actually on the spot can judge what is safe better than legislators and bureaucrats who aren't even there, and who have plenty of incentives to distort the judgment anyway.

you say that with limits, people game the system

When I talked about gaming, I wasn't talking about people trying to avoid being caught speeding. I was talking about legislators and bureaucrats distorting the data about risk to justify lower speed limits and higher fines, when the real reason for them is just that they need a revenue source.

As far as drivers themselves taking more risk in the absence of speeding laws, evidently you missed the part of my post where I said that people should suffer appropriate punishment if they cause actual harm. That gives people a strong incentive to avoid causing actual harm, i.e., to get better at assessing the risks of the things they do, including driving.

To the extent that we actually allow this incentive to operate in our society, I think it works reasonably well, certainly better than trusting governments to get things like this right. shpxnvz provided some evidence for that. Another piece of evidence is the fact that states with no-fault auto insurance have higher accident rates, indicating that uncoupling people's decisions from unpleasant consequences, even partially, causes them to make worse decisions.

Is it also possible that states with no-fault insurance have people more willing to report accidents?

I wouldn't think so; in fact, if anything, I would expect the opposite. If there is an accident in a state with ordinary (i.e., not no-fault) insurance, one party (the one not at fault) is always going to have an incentive to report it. If there is an accident in a no-fault state, it's possible that neither party will want to report it (because it might cause both of their premiums to go up).

>And you do that, IMO, by attaching punishments to actual harm, not to an increased likelihood of harm (which must always be an estimate only).

If you restrict the basis of decision-making to certainty rather than "estimated likelihood", you would have to throw out pretty much everything mankind has ever learned using the scientific method (where evidence can support, or inrease the likelihood of a theory, but never prove it).

For example, the link between smoking and lung cancer. It's impossible to say of any specific case of lung cancer that it was definitely caused by smoking, but there is enough evidence for increased risk, so that deciding to quit based only on a likelihood is a reasonable thing to do.

>the legislator who drafts the speeding laws and the bureaucrat who sets the speed limits for particular roads can't possibly know as much as the driver

In the case of speed traps, that's irrelevant; the limits there are not abnormally low because the legislators and bureaucrats are using unsound information, but because they intentionally ignore (or don't even bother acquiring) recommendations which indicate that a higher speed would be safe.

If you restrict the basis of decision-making to certainty rather than "estimated likelihood"

That's not what I said. I said that the law should not be able to impose punishment based on likelihood of harm, only based on actual harm. That has nothing to do with whether or not people should take into account likelihood of harm when making their own decisions; of course they should.

the limits there are not abnormally low because the legislators and bureaucrats are using unsound information, but because they intentionally ignore (or don't even bother acquiring) recommendations which indicate that a higher speed would be safe

Which is exactly what I have said elsewhere in these posts. In what you quoted, I was only pointing out that in fact, even if the legislators and bureaucrats were trying to set the "right" speed limits; they can't, because they can't possibly have all the relevant information. So the common belief that the legislators and bureaucrats are "helping to keep the roads safe" by setting speed limits is wrong in two ways: (1) that's not their real motivation; and (2) they're not capable of doing it anyway.

Reply on news.ycombinator.com